本文在给定的Poisson样本X1,X2,…,Xn下,研究了Poisson分布变异系数θ的Bayes估计问题,在p,q对称损失函数L(θ,δ)=(θ/δ)p+(δ/θ)q-2(p,q∈Z+),得到了θ的Bayes估计的精确形式并讨论了它的可容许性,最后研究了θ的最大后验区间估计.
In this paper, we dealt with the Bayesian estimation problem for Karl-Pearson coefficient of variance about Poisson distribution, given the Poisson random sample X1, X2, ……, Xn, we obtained the exact form of Bayes estimator and discussed the admissibility of it, using the p, q symmetric loss L(θ,δ)=(θ/δ)p+(δ/θ)q-2,p,q∈Z^+, . Finally, we investigated the maximal posterior interval estimation of coefficient of variance about Poisson distribution.